GPT-5.2 vs Nemotron 3 Ultra
GPT-5.2 leads Nemotron 3 Ultra on coding, 55.9 to 30.0 on the Vibe Coding Index, and Nemotron 3 Ultra costs 5.8x less per output token ($2.40 vs $14.00 per million). Updated October 22, 2026; 60 benchmarks have a result for both.
GPT-5.2 or Nemotron 3 Ultra: which to pick
Pick GPT-5.2 for coding: it scores 55.9 on the Vibe Coding Index against 30.0 for Nemotron 3 Ultra.
Pick Nemotron 3 Ultra for output tokens 5.8x cheaper ($2.40 vs $14.00 per million) and work measured by LiveCodeBench (Vals AI) or LMArena Text: coding, where it leads.
- Shared results
- 60
- GPT-5.2 only
- 56
- Nemotron 3 Ultra only
- 11
- Like-for-like categories
- 8 of 9
Only benchmarks that tested both models count toward a pick; a result only one model has is shown but not compared.
Which one for your work
Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.
Coding
Vibe Coding Index, six public coding benchmarks
GPT-5.2
GPT-5.2 scores 55.9 against 30.0.
3 benchmarks behind each score
Intelligence
AA Intelligence Index
GPT-5.2
GPT-5.2 leads, 30.4 against 22.9.
Same board, both measured
Agentic work
AA Agentic Index
No call
Only Nemotron 3 Ultra has a AA Agentic Index result, 20.1.
Directional only
Reasoning
GPQA Diamond
GPT-5.2
GPT-5.2 leads, 91.7% against 86.1%.
Same board, both measured
Math
AIME
No call
Only GPT-5.2 has a AIME result, 96.9%.
Directional only
Knowledge
MMLU Pro
GPT-5.2
GPT-5.2 leads, 86.2% against 85.8%.
Same board, both measured
Web design
Design Arena: websites
GPT-5.2
GPT-5.2 leads, 1195 against 1145.
Same board, both measured
Long documents
Listed context window
GPT-5.2
GPT-5.2 takes 400K tokens against 262K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Nemotron 3 Ultra
GPT-5.2 $0.0088, Nemotron 3 Ultra $0.0018 per turn.
List prices
Cost per solved task
One SWE-bench Verified task, solved
No call
Only GPT-5.2 has the number, $1.91.
Directional only
At a glance
- Coding index
- GPT-5.2 55.9 from 3 benchmarks, Nemotron 3 Ultra 30.0 from 3 benchmarks.
- Widest category
- Coding, 14.7 points apart on average over 7 shared benchmarks.
- Widest single gap
- Vibe Code Bench (Vals AI): GPT-5.2 53.5% against Nemotron 3 Ultra 7.6%, 45.9 points apart on a 0 to 100 scale.
- Output price
- GPT-5.2 $14.00 against Nemotron 3 Ultra $2.40 per million output tokens, 5.8x apart.
- Context window
- GPT-5.2 takes 400K tokens, Nemotron 3 Ultra 262K.
Benchmark by benchmark
First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.
| Category | GPT-5.2 | Bars | Nemotron 3 Ultra | Ties |
|---|---|---|---|---|
| Coding | 4 | 3 | ||
| Intelligence | 2 | 0 | ||
| Agentic work | 2 | 0 | ||
| Reasoning | 5 | 3 | ||
| Math | 2 | 2 | ||
| Knowledge | 5 | 4 | ||
| Web design | 7 | 1 | ||
| General | 5 | 15 | ||
| All shared boards | 32 | 28 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index43
| Benchmark | GPT-5.2 | Nemotron 3 Ultra | Leader |
|---|---|---|---|
| SWE-bench Verified | 72.8% | Not measured | Only GPT-5.2 measured |
| SWE-bench Multilingual | 66.7% | Not measured | Only GPT-5.2 measured |
| LMArena WebDev | 1417 | Not measured | Only GPT-5.2 measured |
| AA Coding Index | Not measured | 49.3 | Only Nemotron 3 Ultra measured |
| LiveCodeBench (Vals AI) | 85.4% | 86.0% | Nemotron 3 Ultra |
| SWE-bench Verified (Vals AI) | 75.8% | 69.0% | GPT-5.2 |
| Vibe Code Bench (Vals AI) | 53.5% | 7.6% | GPT-5.2 |
| Code migration (Vals AI) | Not measured | 4.9% | Only Nemotron 3 Ultra measured |
| ProgramBench (Vals AI) | Not measured | 0.0% | Only Nemotron 3 Ultra measured |
| Terminal-Bench 2.1 (Vals AI) | Not measured | 50.9% | Only Nemotron 3 Ultra measured |
| LiveCodeBench (Artificial Analysis run) | 89.4% | Not measured | Only GPT-5.2 measured |
| SciCode | Not measured | 40.3% | Only Nemotron 3 Ultra measured |
| Terminal-Bench Hard (Artificial Analysis run) | 47.0% | 36.4% | GPT-5.2 |
| Terminal-Bench | 64.9% | Not measured | Only GPT-5.2 measured |
| FrontierCode | Not measured | 13.6% | Only Nemotron 3 Ultra measured |
| GSO-Bench | 27.5% | Not measured | Only GPT-5.2 measured |
| WeirdML | 72.2% | 43.5% | GPT-5.2 |
| ALE-Bench | 1293.5 | Not measured | Only GPT-5.2 measured |
| LMArena Text: coding | 1442 | 1468 | Nemotron 3 Ultra |
| LMArena Text: software and IT | 1434 | 1468 | Nemotron 3 Ultra |
| LMArena WebDev: HTML | 1427 | Not measured | Only GPT-5.2 measured |
| Neither model has a result on the other 22 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index20
| Benchmark | GPT-5.2 | Nemotron 3 Ultra | Leader |
|---|---|---|---|
| AA Intelligence Index | 30.4 | 22.9 | GPT-5.2 |
| Epoch Capabilities Index | 153.7 | 146.0 | GPT-5.2 |
Agentic workLike-for-likePublic boards, shown as published, kept out of the index20
| Benchmark | GPT-5.2 | Nemotron 3 Ultra | Leader |
|---|---|---|---|
| AA Agentic Index | Not measured | 20.1 | Only Nemotron 3 Ultra measured |
| Design Arena: agents building web apps | 1086 | Not measured | Only GPT-5.2 measured |
| Design Arena: agents building full-stack apps | 1043 | Not measured | Only GPT-5.2 measured |
| Design Arena: agents building mobile apps | 1093 | Not measured | Only GPT-5.2 measured |
| Design Arena: agents building Android apps | 1059 | Not measured | Only GPT-5.2 measured |
| Design Arena: agents building Godot games | 1142 | Not measured | Only GPT-5.2 measured |
| Finance Agent (Vals AI) | Not measured | 37.7% | Only Nemotron 3 Ultra measured |
| Legal Research Bench (Vals AI) | Not measured | 15.4% | Only Nemotron 3 Ultra measured |
| Excel Modeling Benchmark (Vals AI) | Not measured | 31.7% | Only Nemotron 3 Ultra measured |
| τ²-Bench Telecom | 84.8% | 83.3% | GPT-5.2 |
| APEX-Agents | Not measured | 22.7% | Only Nemotron 3 Ultra measured |
| METR time horizons | 75.3% | Not measured | Only GPT-5.2 measured |
| Remote Labor Index | 2.5% | Not measured | Only GPT-5.2 measured |
| GDPval | 49.7% | Not measured | Only GPT-5.2 measured |
| DeepResearch Bench | 41.1% | Not measured | Only GPT-5.2 measured |
| Vending-Bench 2 | $3,591 | Not measured | Only GPT-5.2 measured |
| Berkeley Function Calling Leaderboard (BFCL) | 55.9% | Not measured | Only GPT-5.2 measured |
| MCP Atlas | 67.6% | 63.1% | GPT-5.2 |
| Neither model has a result on the other 24 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index53
| Benchmark | GPT-5.2 | Nemotron 3 Ultra | Leader |
|---|---|---|---|
| GPQA Diamond | 91.7% | 86.1% | GPT-5.2 |
| MMMU Pro (Vals AI) | 86.7% | Not measured | Only GPT-5.2 measured |
| GPQA Diamond (Artificial Analysis run) | 90.3% | 86.7% | GPT-5.2 |
| Humanity's Last Exam (Artificial Analysis run) | 37.7% | 28.4% | GPT-5.2 |
| GPQA Diamond (Epoch AI run) | 91.4% | 85.4% | GPT-5.2 |
| ARC-AGI | 86.2% | Not measured | Only GPT-5.2 measured |
| ARC-AGI-2 | 52.9% | Not measured | Only GPT-5.2 measured |
| SimpleBench | 45.8% | Not measured | Only GPT-5.2 measured |
| Chess puzzles | 48.0% | Not measured | Only GPT-5.2 measured |
| Mystery game puzzles | 23.0% | 20.0% | GPT-5.2 |
| EBR-bench | 23.0% | Not measured | Only GPT-5.2 measured |
| VPCT | 84.0% | Not measured | Only GPT-5.2 measured |
| LMArena Text: hard prompts | 1428 | 1452 | Nemotron 3 Ultra |
| LMArena Text: expert | 1445 | 1472 | Nemotron 3 Ultra |
| LMArena Text: hard prompts in English | 1431 | 1453 | Nemotron 3 Ultra |
| LMArena Vision: diagrams | 1265 | Not measured | Only GPT-5.2 measured |
| LMArena Vision: homework | 1276 | Not measured | Only GPT-5.2 measured |
| Kagi LLM Benchmark | 73.3% | Not measured | Only GPT-5.2 measured |
| Neither model has a result on the other 5 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index22
| Benchmark | GPT-5.2 | Nemotron 3 Ultra | Leader |
|---|---|---|---|
| AIME | 96.9% | Not measured | Only GPT-5.2 measured |
| MGSM (Vals AI) | 94.0% | Not measured | Only GPT-5.2 measured |
| AA Math Index | 99.0 | Not measured | Only GPT-5.2 measured |
| AIME 2025 (Artificial Analysis run) | 99.0% | Not measured | Only GPT-5.2 measured |
| OTIS Mock AIME | 96.1% | 86.7% | GPT-5.2 |
| FrontierMath, tiers 1 to 3 | 67.4% | Not measured | Only GPT-5.2 measured |
| FrontierMath, tier 4 | 31.7% | Not measured | Only GPT-5.2 measured |
| ProofBench | 15.0% | 2.0% | GPT-5.2 |
| LMArena Text: math | 1440 | 1457 | Nemotron 3 Ultra |
| LMArena Text: mathematical work | 1442 | 1461 | Nemotron 3 Ultra |
| MathArena | 33.7% | Not measured | Only GPT-5.2 measured |
| Neither model has a result on the other 3 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index54
| Benchmark | GPT-5.2 | Nemotron 3 Ultra | Leader |
|---|---|---|---|
| MMLU Pro | 86.2% | 85.8% | GPT-5.2 |
| LegalBench (Vals AI) | 82.8% | 82.1% | GPT-5.2 |
| CaseLaw (Vals AI) | 66.0% | Not measured | Only GPT-5.2 measured |
| CorpFin (Vals AI) | 65.9% | 65.5% | GPT-5.2 |
| TaxEval (Vals AI) | 75.8% | 73.1% | GPT-5.2 |
| MortgageTax (Vals AI) | 67.1% | Not measured | Only GPT-5.2 measured |
| MedQA (Vals AI) | 94.1% | Not measured | Only GPT-5.2 measured |
| MedCode (Vals AI) | 49.8% | 38.6% | GPT-5.2 |
| MMLU Pro (Artificial Analysis run) | 87.4% | Not measured | Only GPT-5.2 measured |
| SimpleQA Verified | 37.1% | Not measured | Only GPT-5.2 measured |
| LMArena Text: science | 1425 | 1465 | Nemotron 3 Ultra |
| LMArena Text: medicine and healthcare | 1425 | 1456 | Nemotron 3 Ultra |
| LMArena Text: legal and government | 1427 | 1452 | Nemotron 3 Ultra |
| LMArena Text: business and finance | 1414 | 1434 | Nemotron 3 Ultra |
| FACTS Benchmark Suite | 48.5% | Not measured | Only GPT-5.2 measured |
| Neither model has a result on the other 5 boards here | |||
Web designLike-for-likePublic boards, shown as published, kept out of the index71
| Benchmark | GPT-5.2 | Nemotron 3 Ultra | Leader |
|---|---|---|---|
| Design Arena: websites | 1195 | 1145 | GPT-5.2 |
| Design Arena: UI components | 1188 | 1135 | GPT-5.2 |
| Design Arena: game dev | 1195 | 1138 | GPT-5.2 |
| Design Arena: data viz | 1202 | 1141 | GPT-5.2 |
| Design Arena: 3D | 1079 | 1131 | Nemotron 3 Ultra |
| Design Arena: all code categories | 1170 | 1146 | GPT-5.2 |
| Design Arena: SVG | 1139 | 1069 | GPT-5.2 |
| Design Arena: ASCII art | 1152 | 1100 | GPT-5.2 |
| Neither model has a result on the other 1 board here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index515
| Benchmark | GPT-5.2 | Nemotron 3 Ultra | Leader |
|---|---|---|---|
| SAGE (Vals AI) | 49.3% | Not measured | Only GPT-5.2 measured |
| MedScribe (Vals AI) | 84.4% | Not measured | Only GPT-5.2 measured |
| IFBench | 75.4% | 81.4% | Nemotron 3 Ultra |
| Long context reasoning (AA-LCR) | 82.7% | 79.3% | GPT-5.2 |
| DTBench | 90.9% | 90.1% | GPT-5.2 |
| LMCA | 43.9% | 36.9% | GPT-5.2 |
| CL-bench | 18.2% | Not measured | Only GPT-5.2 measured |
| Furniture assembly | 38.3% | Not measured | Only GPT-5.2 measured |
| LMArena Text | 1416 | 1445 | Nemotron 3 Ultra |
| LMArena Text: instruction following | 1409 | 1418 | Nemotron 3 Ultra |
| LMArena Text: creative writing | 1376 | 1399 | Nemotron 3 Ultra |
| LMArena Vision | 1246 | Not measured | Only GPT-5.2 measured |
| LMArena Text: English | 1422 | 1455 | Nemotron 3 Ultra |
| LMArena Text: non-English | 1405 | 1427 | Nemotron 3 Ultra |
| LMArena Text: Chinese | 1447 | 1497 | Nemotron 3 Ultra |
| LMArena Text: German | 1444 | 1472 | Nemotron 3 Ultra |
| LMArena Text: Japanese | 1420 | Not measured | Only GPT-5.2 measured |
| LMArena Text: Korean | 1374 | 1384 | Nemotron 3 Ultra |
| LMArena Text: French | 1454 | 1472 | Nemotron 3 Ultra |
| LMArena Text: Spanish | 1413 | 1455 | Nemotron 3 Ultra |
| LMArena Text: Russian | 1415 | 1417 | Nemotron 3 Ultra |
| LMArena Text: Polish | 1426 | 1385 | GPT-5.2 |
| LMArena Text: multi-turn | 1419 | 1411 | GPT-5.2 |
| LMArena Text: longer queries | 1413 | 1437 | Nemotron 3 Ultra |
| LMArena Text: writing and language | 1392 | 1406 | Nemotron 3 Ultra |
| LMArena Text: entertainment and media | 1378 | 1397 | Nemotron 3 Ultra |
| LMArena Vision: English | 1243 | Not measured | Only GPT-5.2 measured |
| LMArena Vision: Chinese | 1305 | Not measured | Only GPT-5.2 measured |
| LMArena Vision: reading text | 1257 | Not measured | Only GPT-5.2 measured |
| LMArena Vision: humor | 1227 | Not measured | Only GPT-5.2 measured |
| LMArena Vision: captioning | 1267 | Not measured | Only GPT-5.2 measured |
| LMArena Vision: recognizing things | 1232 | Not measured | Only GPT-5.2 measured |
| LMArena Vision: creative writing | 1225 | Not measured | Only GPT-5.2 measured |
| LMArena Document | 1405 | Not measured | Only GPT-5.2 measured |
| Neither model has a result on the other 14 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | GPT-5.2 | Nemotron 3 Ultra | Leader |
|---|---|---|---|
| Neither model has a result on 5 boards here | |||
What each workload costs
Three fixed token mixes at each model's listed price, so per-token rates become a bill.
Chat turn
1K input + 500 output tokens
Nemotron 3 Ultra costs less
Repository review
50K input + 3K output tokens
Nemotron 3 Ultra costs less
Agent loop
200K input + 10K output tokens
Nemotron 3 Ultra costs less
Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.
Specification differences
The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.
| Specification | OpenAI GPT-5.2 | NVIDIA Nemotron 3 Ultra |
|---|---|---|
| Maker | OpenAI | NVIDIA |
| Weights | No public weights listed | Open weights |
| Released | December 10, 2025 | June 4, 2026 |
| Context window | 400K | 262K |
| Input $/Mtok | $1.75 | $0.60 |
| Output $/Mtok | $14.00 | $2.40 |
| Coding index | 55.9 | 30.0 |
| Independent measurements behind it | 3 | 3 |
| Cost per solved task | $1.91 | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is GPT-5.2 better than Nemotron 3 Ultra for coding?
Pick GPT-5.2 for coding: it scores 55.9 on the Vibe Coding Index against 30.0 for Nemotron 3 Ultra. Pick Nemotron 3 Ultra for output tokens 5.8x cheaper ($2.40 vs $14.00 per million) and work measured by LiveCodeBench (Vals AI) or LMArena Text: coding, where it leads.
Which is cheaper, GPT-5.2 or Nemotron 3 Ultra?
GPT-5.2 costs $1.75 per million input tokens and $14.00 per million output tokens; Nemotron 3 Ultra costs $0.60 and $2.40. Cost per solved task is known for GPT-5.2 only.
Which has the bigger context window, GPT-5.2 or Nemotron 3 Ultra?
GPT-5.2, with 400K tokens against 262K.
How many benchmarks is the GPT-5.2 vs Nemotron 3 Ultra comparison based on?
60 benchmarks have a result for both models as of October 22, 2026. GPT-5.2's coding index rests on 3 independent benchmarks and Nemotron 3 Ultra's on 3. Where only one model has a result, the page says so instead of naming a winner.